An Optimal DoS Attack Strategy With Pause and Restart Rules Under Energy Constraints

This study investigates a denial-of-service (DoS) attack strategy against cyber-physical systems (CPS) under energy constraints. A DoS attacker with eavesdropping ability under a limited energy budget was considered. By scheduling the attack behavior based on historical attack results, the attacker aims to maximize the state estimation error of the CPS. A novel DoS attack strategy with pause and restart rules was proposed, in which the attacker pauses the attack for a while to restore energy after a failed attack attempt. The restart rule was described based on Markov chain modeling of the attacking period. The transition matrix and stationary distribution of the Markov chain model were also presented. To maximize the attack effect, the optimal restart rule of the proposed strategy was investigated. Finally, it was proven that the proposed attack strategy in this study is more destructive than the periodic attack strategy, and simulations were provided to verify our theoretical results.